Estimating a ratio of normal parameters
Let \(x_ 1,...,x_ n\) (n\(\geq 2)\) be a sample from a normal distribution with mean \(\xi\) and variance \(\sigma^ 2\). The parameter \(\theta =\xi /\sigma^{2p}\) is to be estimated with loss function \(L(\xi,\sigma,\delta)=(\theta -\delta)^ 2\sigma^{4p-2}\). Consider estimators \(CX/S^ p\) where X is the sample mean and \(S=\sum^{n}_{j=1}(x_ j-\bar X)^ 2\). The unbiased estimator, given by \[ C=2^ p\Gamma ((n-1)/2)/\Gamma ((n-1)/2-p), \] can be improved upon when \(p\neq 0\), \(n>4p+1\) by setting \[ C=2^ p\Gamma ((n-1)/2- p)/\Gamma ((n-1)/2-2p). \] If \(p\geq 0\) and \(n>4p+1\), the latter estimator is admissible.
- Quadratic estimators of quadratic functions of normal parameters
- Generalized Bayes estimators of a normal discriminant function
- Ratios of linear combinations in the general linear model
- A sequential procedure for estimating ratio of normal parameters
- scientific article; zbMATH DE number 503312 (Why is no real title available?)
- On the estimation of a normal precision and a normal variance ratio
- Sequential estimation of ratio of normal parameters
This page was built for publication: Estimating a ratio of normal parameters
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1074265)
